Session 3: Chamber Armor (PFC) and Blanket Tritium Inventory and Recovery

Chairman: A. R. Raffray

Presentations:

1. Tritium inventory and recovery experience in tokamaks (C. H. Skinner)
2. Tritium issues for PFC (R. Causey)
3. Tritium retention issues in the ITER-FEAT device (G. Federici)

4. Tritium breeding issues for MFE and IFE (L. El-Guebaly)
5. Tritium recovery and confinement from breeding blanket (D. Sze)
6. Discussion and summary (R. Raffray)

The first three presentations (1-3) covered tritium inventory and recovery associated with different plasma facing materials based on experience from existing machines, and from small-scale experimental results and modeling analysis. The key points from these presentations are summarized below for different plasma facing component (PFC) materials.

Carbon

• At temperatures lower than 800 K in vacuum, hydrogen implanted in carbon is immobile
• Tritium can be implanted in carbon until a saturated layer is formed with a H/C ratio of ~0.4 at room temperature.
• Tritium co-deposition with carbon is the major tritium retention process for PFC applications with inventory corresponding to a H/C ratio of up to 1.
• Large uncertainties still exist in computer codes for estimating carbon erosion and tritium co-deposition. The code used has not been validated for detached-plasma C erosion
• Within these uncertainties, application of these codes to estimate erosion and tritium co-deposition inventory in carbon PFC during ITER operation indicates gross and net peak divertor erosion of about 65 and 6 nm/s, respectively. This is a source for concern as it would lead to large tritium co-deposition rates and the need for frequent in-situ recovery of this tritium, and necessitate frequent replacement of the divertor (e.g., in ITER-FEAT after less than 3000 discharges if other mechanisms such as disruptions are included).
• Even given the uncertainties in code predictions, it is clear on the basis of modeling analysis and experimental observation that tritium retention and erosion in carbon are unsustainable in MFE power plant reactors.
• Neutron effects on carbon properties, in particular thermo-mechanical properties, can be important and must be taken into account when designing and analyzing carbon PFC¹s. This is particularly important for low-temperature operation such as for water-cooled components since the relative radiation effect tends to increase with decreasing temperature.

Tungsten

• There is a much lower tritium retention in tungsten than in C with tungsten having the advantage of not co-depositing with tritium.
• Tungsten has a high threshold for sputtering and would be quite resistant to erosion in this regard.
• A tungsten brush configuration appears to offer good accommodation of surface heat fluxes and a test sample tested at SNL accommodated a heat flux of up to ~20 MW/m2 without any discernable damage. Comparable and even better results have also been obtained in Russia and the EU.
• Accommodation of high energy deposition under transient off-normal conditions such as disruptions and plasma vertical displacement events (VDE¹s) is a concern because of the possible formation of a melt layer. It is not clear whether such a layer will resolidified or will be partially lost which would create a lifetime issue. In addition, the lifetime will be affected by the possible formation of blisters under such conditions. Another concern relates to the possible formation of hot spots during normal operation arising from surface irregularities caused by off-normal events.
• Scarce data exist on neutron effects on tungsten properties and R&D effort is needed in this area.
• Another concern with tungsten is its high activation and decay heat which must be considered in safety analysis scenarios.

Beryllium

• The tritium solubility in Be is very low.
• A tritium saturated layer is formed at the surface when tritium is implanted on a Be PFC surface or when Be is exposed to tritium.
• No tritium co-deposition has been observed in Be. However, co-deposition is possible with BeO.
• For tritium bred in Be, 100% retention is expected at low temperatures (~40°C) but tritium will be released at higher temperature
• Charge exchange erosion of Be is a concern. For ITER, Be erosion of up to 0.1 nm/s has been estimated. This would be acceptable for a low duty factor reactor such as ITER but would be unacceptable for power plant reactors.
• Safety concerns exist with Be, not only regarding dust and handling, but more importantly the potential reaction of Be with high temperature steam under accident scenarios and the production of hydrogen (which, in the presence of air, can lead to explosions).
• Neutron-induced swelling causes microcracking of Be even al low fluences.

A few observations and comments were also made about other potential PFC materials.

• Vanadium takes up and stores tritium which would come out at high temperature.
• Up to 100% tritium retention has been observed in Li up to ~700°C (until saturation is reached). However, recent PISCES data indicate much lower retention in Li (~1%). These results are not yet fully understood and are still being analyzed. The solubility of hydrogen species is about 3 orders of magnitude higher in Li than it is in tin. Hydrogen species also have a very high diffusivity in Li.
• The PFC behavior of SiC composites is probably very similar to that of C composites, even for tritium co-deposition.

Key remaining R&D issues that bear on the selection of plasma facing components include:

• To better understand the behavior of mixed materials (e.g. a combination of C, W and/or Be in ITER).
• To better understand and to accommodate the effect of disruption and edge-localized mode (ELM) loads on PFC thermo-mechanical behavior, including the melt layer behavior for metal armor, vapor shielding effect, and the possibility of brittle destruction.
• To fully characterize the effect on erosion (for Be and possibly for W) of advanced plasma scenario with high edge temperatures.
• The necessity of developing and implementing in-vessel dust diagnosis to demonstrate safety limit compliance.
• Understanding plasma flows in the scrape off region to better predict the amount and location of co-deposited tritium.
• Eliminating or mitigating off-normal high heat flux events such as ELM¹s and disruptions to permit more flexibility in material choice.
• The need to define the damage limits for C, W, and Be under 14 MeV neutron radiation.

The final two presentations (4-5) covered tritium breeding and blanket tritium inventory issues.

Breeding Issues

• There is confidence that tritium self-sufficiency can be achieved.
• A margin of ~10% in estimating the tritium breeding ratio is required to account for uncertainties in cross-section data and geometric modeling, and for tritium supply for new power plants, and decay/losses.
• The blanket design should offer solutions to adjust TBR, preferably in-situ, for example through 6Li enrichment adjustment in the case of a flowing liquid breeder. Other means of adjusting breeding requiring hardware replacement such as adjusting breeding region thicknesses, volume fractions, and Be content are possible but less attractive. Overall, it is preferable to have overbreeding than underbreeding.
• It seems easier to achieve the required tritium breeding in IFE than in MFE as there are fewer constraints on breeding region thickness and volumetric space loss due to penetrations.

Blanket Tritium Inventory

• Tritium can be recovered from breeding materials at the steady state production rate with an inventory of < 10 g for ceramic breeders, ¾ 10 g for Flibe, and < 200 g for Li (~ 1 appm)
• Keeping tritum leakage rate < 1-10 Ci per full-power-day is a major concern. The primary heat exchanger is the main tritium leakage window and a partial pressure of <10-7 Pa is required to limit leakage to 10 Ci/FPD. However, the tritium partial pressure in the different fluids considered is much higher (except for Li):
  • 10 Pa in Flibe;
  • 1 Pa in Pb-17Li
  • 10^-9 Pa in Li; and
  • 200 Pa in He (used as purge gas for ceramic breeder blanket)
• Thus, the partial pressure in each of these cases (except for Li) must be reduced by orders of magnitude. Possible methods to help solve this key issue include:
  • Efficient recovery
  • Double-wall heat exchanger; permeation barrier; and/ or low permeation structural material
  • Secondary recovery of tritium
  • Use power conversion without H-containing fluids (e.g. He)

Discussion

The discussion session focused on comparing chamber wall material behavior operating conditions and behavior for MFE and IFE applications. Since IFE is based on pulsed depositions of photon and ion energies on the chamber wall (with frequency of ~5-10 Hz), it seems interesting to compare the information obtained from MFE PFC's under transient (usually off-normal) conditions to see if they can be applied to help understand the IFE chamber armor behavior. A table summarizing the conditions assumed for ITER ELM's, VDE's and disruptions and the conditions for a typical direct drive target IFE was evolved and is shown in Table 1. In general, the energy deposition density, the time scale and the frequencies for the MFE cases are quite different than those for IFE. However, for the case of ELM's there are some interesting similarities as the frequency is roughly the same as for the IFE case, and the energy deposition density and time scale are within about one and two orders of magnitude, respectively, of the corresponding values for the IFE case. Although, one should be very cautious in applying MFE results on wall behavior to the IFE case, there is a wealth of information from MFE PFC R&D which might well be of relevance for the IFE case. It would seem wise for IFE researchers to make the most of this and to bring in the expertise of the MFE PFC and material community when pursuing IFE chamber armor design and analysis.

From the ensuing discussion on IFE expectations based on MFE experience, the following points emerged:

• Carbon erosion could lead to tritium co-deposition, raising both tritium inventory and lifetime issues for IFE with a carbon wall. Redeposition/co-deposition requires cold surfaces which would exist in the beam penetration lines and pumping ducts.
• Macroscopic erosion might be a more important lifetime issue than sputtering and sublimation for IFE operating conditions for high energy ions (>>1 keV)
• Overall, the required R&D effort for IFE armor material should not be underestimated